「経験の第二の類推」についての覚書

Abstract

This paper is to give, in a somewhat informal manner, some critical reexaminations on so called Kant's reply to Hume: Kant's proof of the general principle of causality, contained in 'Second Analogy' of the first Critique. First, the exact place that section has in the whole Transcendental Analytic is explicated. In my view, the objective validity of the Categories are fully established in the Transcendental Deduction; after this being done, the Second Analogy (like all the other sections in 'the System of all the Principles of the pure Understanding') is concerned solely with justifying the theory of the Schematism of the Categories, which is to explain how the latter can be applied to phenomena. The justification is done by showing that experience (empirical time-determination) is only possible by means of some transcendental schema; showing this comes down to confirming that the transcendental schemata as they are formulated by Kant are really the applied Categories derived from the forms of judgment. Secondly, an attempt at proper reconstruction of the argument in the Second Analogy is made. Its starting premise, which is a result of the Transcendental Deduction, is that it is possible to determine empirically objective temporal relations of phenomena. But it depends heavily upon some further presuppositions, 1) those concerning the nature of our 'perceptions' and 2) those concerning the notion of 'cause' or 'causality'. To begin with, by 1), Kant denies the perceptivility of temporal relations between any two phenomena either vis-à-vis the absolute time or vis-à-vis one another. This means that no successive sequence of perceptual representations, however long, could provide us with any information about objective temporal relations among phenomena. It follows from this that whenever an objective succession of phenomena is experienced (i.e. the objective temporal relation between them is empirically determined as succesive), some (particular inferential) rule concerning their time-relation must be presupposed to obtain in that situation. Then, from this, Kant infers that such a rule must be a particular causal law, according to 2): i.e. that every non-logical necessary connection is a causal one, that cause cannot be preceded by effect and that causal relation is asymmetrical. As a result of this reconstrucion, it is shown that Kant's argument is logically sound, if only those presuppositions mentioned above be accepted. Thirdly, some important implications of the argument of the Second Analogy are pointed out. The first point is that this argument shows nothing about cognizability of those particular causal laws presupposed to obtain in the objects themselves. That is, the objective validity of the general principle of causality does not imply at all that there is any amount of regularity or order shown in nature. The second point is that the Second Analogy establishes that the principle of causality is one of those formal principles belonging to the judging subject alone (i.e. 'thrown in' or 'imposed' a priori by this subject). This is obvious from the place of the Analogies in the Transcendental Analytic explained above. (Because of lack of space, I should like to discuss on another occasion whether and why those presuppositions of Kant's argument pointed out above are unacceptable.)